This competing renewal application proposal has 4 specific aims and will be conducted by a multidisciplinary team over 4 years. Aim 1 is to analyze patterns of alcohol consumption over the life course, using 6 waves (1985-2000) longitudinal data from at least the first 7 waves of the Health and Retirement Study (HRS) for persons aged 50+ and from Coronary Artery Risk Development in Young Adults (CARDIA) for persons younger than this. Our goal is to compare effects of determinants of drinking, e.g., stressful life events, education, and psychological traits in young adulthood and in late middle through old age, with the end result being an integrated model of alcohol consumption over the life course. In Aim 2, we assess effects on occupational attainment and on labor market outcomes. Past research has suggested that light/moderate drinkers experience more favorable outcomes than either abstainers or heavy drinkers. This may reflect endogeneity of alcohol consumption and/or confounding (failure to account for variables systematically correlated with drinking patterns). We will use the panel structure of CARDIA and factors not included in previous research on alcohol consumption effects (e.g., psychological traits) to deal with confounding in occupational choice. In Aim 3, we address 2 issues: (1) How do market size and regulation affect competitiveness and number of on-site (bars, restaurants) and off-site alcohol sellers (liquor stores)?; (2) How is the spatial distribution of such sellers determined? We will use data from Economic Census and other Census data for this aim. In Aim 4, we examine the effect of availability of alcohol sellers, measured by density of sellers near the person's place of residence, on alcohol consumption of respondents to CARDIA and HRS. Since the number of sellers and place of residence may be endogenous, an important part of Aim 4 involves investigating the role of alcohol availability in the person's location decision and the endogeneity or seller location, largely drawn from Aim 3 and identifying effects of alcohol availability on consumption from data on nonmovers if endogeneity proves to be important.